Search for a command to run...
Introduction: Delay of detection and treatment of nonconvulsive seizures (NCS) is associated with poor outcomes, and timely electroencephalography (EEG) monitoring is necessary for their detection. AHA guidelines recommend EEG monitoring in patients who achieve return to spontaneous circulation (ROSC) after cardiac arrest, however access to EEG remains limited. We pose that using point-of-care (POC) EEG with artificial intelligence (AI) algorithms in critical care will fill accessibility gaps and impact patients’ outcomes. Methods: Subanalysis of retrospective data of adult patients at three community Mercy MO hospitals who received a ≥30-min EEG as part of their standard of care, within 7 days of admission. Sites had limited conventional EEG (convEEG) capability prior to POC EEG adoption. Two cohorts were enrolled: patients monitored with convEEG (Jun-Dec 2019) and those monitored with POC EEG after its adoption (Jun-Dec 2023). We included only patients’ first eligible visit, and those with cardiac arrest as admission diagnosis or EEG indication. We analyzed the time from first contact to EEG start (door-to-EEG) and modified Rankin Scale (mRS) score at discharge. We extracted EEG findings neurology reports and output from the AI algorithm accessible at the bedside. Results: Thirty-two patients were included (convEEG = 12, POC EEG = 20). Mean age (64y (15.8) vs. 61y (10.8); p = 0.8) and median GCS score on admission (5 [3,14] vs. 8 [3,15]; p = 0.59) were comparable between cohorts. Median door-to-EEG time was significantly shorter for POC EEG (26.0 vs. 7.2 h, p< 0.001). Of patients presenting to the hospital afterhours (weekends or weekdays 5pm-8am), a greater proportion of those receiving POC EEG vs. convEEG started their EEG afterhours (0% vs. 90%, p = 0.002). In the convEEG cohort, no seizures were noted, while in the POC EEG arm, the AI algorithm detected continuous seizure activity in one case and was treated after neurology review. At discharge, the mortality rate was similar between cohorts (67% vs. 55%, p = 0.70), and a higher rate of POC EEG cases had favorable mRS scores ≤ 3 (0% vs. 35%; p = 0.03). Conclusions: POC EEG offered enhanced coverage in the afterhours and facilitated management of NCS in post-ROSC patients. As a group, the POC EEG patients had better functional outcomes at discharge.